Modern graphic processors perform a number of events in parallel. For example, a graphics processors can perform combinations of different events, such as, clear, copy, draw, dispatch, or the like in parallel. The order of event processing is constrained based on various dependencies between the events and/or between the various frames to which the events correspond. Due to the highly parallel nature and due to the constraints of events processing, identifying areas where inefficiencies are introduced can be a challenge.